The Art and Science of Diagnosing

[Admin]

Diagnosis is (wrongly) held to be the central process of clinical medicine.
We submit that the central processes of medicine are: relieving symptoms, providing reassurance and prognostic information, and lending a sympathetic ear. But it is very difficult to do this well, and to increase your rapport with our patient, unless you have a working diagnosis. How is this achieved?

We diagnose, it is held, by a 3-stage process: we take a history, we examine and we do tests. We then collate this information, by a process which in never explained, and compare it with features of disease we know.
We then find the best match, and call this the diagnosis. Other nearly matching diseases then form the differential diagnosis (Known as D/D). This model ignores two factors:
1- Often no match can be found
2- Doctors, in practice, hardly ever work like this. So how are diagnoses made?!!

Diagnosing by recognition: For Students, this is the most irritating method.
You spend an hour asking all the wrong questions, and in waltz a doctor who names the disease, and sorts it out before you have even finishied taking the pulse!. This doctor has simply recognized the illness like he recognizes an old friend (or enemy). But don't worry: you too will soon reach this position. if you spend enough time at bedsides with other doctors-and you too will just as effortssly make all the errors that this approach is prone to.

Diagnosing by probability theory: Over our clinical lives we unconsciously build up a personal database of diagnoses and outcomes, and associated pitfalls. We unconsciously run each new 'case' through this personal and continuously developing fine-grained probabilistic algorithm-eventually with amazing speed and effortlessness.

Diagnosing by reasoning: Like Sherlock Holmes, we exclude each differential diagnosis, then, whatever is left, however unlikely, must be the culprit. This process presupposes that your differential does include the culprit, and that we have methods for absolutely excluding disease, All tests are statistical, rather than absolute-Which is why the Holmes technique is, at best, fictional.

Continue reading below
Dr.Mahdi

[Admin]

Diagnosing by WoE: WoE stands for 'wait on events'-and the notes of any good diagnostician will be littered with this injunction. Some doctors need to know immediately and definitively what the diagnosis is, while others can tolerate more uncertainty. with proactise, one can sense that patients are not it extremes, and that the dangers and expense of exhaustive tests can bee obviated by the skillful use of time. This cough might represent pneumonia, but I may choose not to prove this now by plating out a pulmonary aspirate in the microbiology department. Rather, I may say 'take this prescription if you get feverish, or your sputum becomes green-but you probably don't you need anything from me, and your body will simply cure itself: wait and see.

Diagnosing by hypothesizing: We formulate a hypothesis and then try to disprove or prove it. Proof is elusive, but, in proactise, who does doubt the blood's circulation, or the notion that humans are vertebrates?

Diagnosing by selective doubting: Traditionally, patients are 'unreable' Signs are objective, and lab results virtually prefect. When diagnosis is difficult, try inverting this hierarchy. The more you do so, the more you realize that there are no hard signs or perfect labs. But the game of medicine is unplayable if you doubt everything: so doubt selectively, bearing in mid Wittgenstein's dictum that unless you can doubt an entity, you can never be said to know it.

Diagnosing by Computer: Computing power is the only way of fully mapping the interrelatedness of disease -eg Na+ decrease with easinophilia points to Addison's disease, but if there is oliguria too, a doctor with no computer may have to posit an unrelated disease; but the computer 'knows' that oliguria is a feature of shock, and shock is a complication of Addison's.

END
From Oxford Handbook of Clinical Medicine
Dr.Mahdi